XML vs AnnotationCollectionI [was Re: [Bioperl-l] AnnotationCollectionI and SeqFeatureI changes]

[Aaron J. Mackey]

I think this is a grand idea for what I've called BioPerl "nouveau", in 
which we may use any and all existing technologies at our disposal, 
regardless of how convoluted and difficult the installation may be, 
runtime-instantiated methods (even classes) are allowed, and where we 
only care marginally about backwards compatibility, and instead focus 
on doing it "right" (of course there are large arguments about this).

In the "nouveau" project, interfaces are hidden from the naive 
end-user, documentation is highly structured (and inheritable), parsing 
is block- and event-based, and there are flexible API "adaptors" to 
provide ultra-simple, light-weight or heavy-weight access to underlying 
objects.  Sounds great, but still a pipe-dream, unfortunately.

I've been (slowly) writing some use cases for bioperl-nouveau which are 
really just example end-user code making use of simple, light-weight 
and heavy-weight APIs.  Since this is still just a hobby for me, I 
haven't made them public yet; is there enough interest (yet) to talk 
about what a next-generation BioPerl might look like?

-Aaron

P.S. But Chris, if you actually wanted to get some real work done, 
you're welcome to just make a branch and go crazy ... whether we end up 
using it in the existing BioPerl depends on whether it ends up being 
more useful than the current implementation, and whether the audience 
seems willing to install XML tools for such a "core" functionality.

On Nov 23, 2004, at 10:24 PM, Chris Mungall wrote:

> I may play around with something like this on a clean branch if there's
> anyone else who doesn't think this is a mad idea and may actually use 
> the
> final results....
--
Aaron J. Mackey, Ph.D.
Dept. of Biology, Goddard 212
University of Pennsylvania       email:  amackey at pcbi.upenn.edu
415 S. University Avenue         office: 215-898-1205
Philadelphia, PA  19104-6017     fax:    215-746-6697